Real-time stress monitoring in a child-friendly court: a repeated measures field study

The study investigates whether children testifying in a child-friendly investigative court experience high, potentially harmful levels of acute stress. Acute stress can adversely affect brain functioning and is implicated in secondary victimization in judicial contexts. Prior findings on the impact of testifying are mixed, with some reporting negative outcomes and others finding no harm or even restorative effects. European child-friendly justice guidelines emphasize adapting procedures, environments, and language to reduce stress. Spain’s judicial system has implemented preconstituted evidence and specialized environments (e.g., Gesell rooms) to protect minors. In 2021, the first Spanish court specialized in violence against children incorporated real-time stress monitoring via a smart band to prevent high acute stress during testimony. The research question asks: Do children endure high and damaging levels of stress in child-friendly courts? The objective is to analyze acute stress levels using a person-by-environment approach, with repeated real-time stress measures across different courtroom contexts.

Evidence suggests courtroom participation may elevate stress and anticipatory anxiety in minors (e.g., Goodman et al., 1998; Saywitz & Nathanson, 1993; Nathanson & Saywitz, 2003), though impacts vary with conditions of participation (Carretta & García-Quiroga, 2021). European and FRA guidelines advocate for child-friendly facilities and procedures to mitigate stress and secondary victimization. Spain expanded videotaped preconstituted evidence from sexual violence cases to all minors under 14, aiming to protect both the child and evidentiary integrity. Modern Gesell rooms and CCTV seek to reduce stressors compared to traditional settings. HRV is a validated biomarker for stress and anxiety-related autonomic regulation, with wearables providing minimally intrusive monitoring. Prior real-time HRV studies with children were largely simulated, showing higher stress in courtroom-like settings versus private rooms (Nathanson & Saywitz, 2003). There is a noted gap in real-time, in-situ measurements with real victims across the testimony process in child-friendly courts.

Design: Simple repeated-measures design with one within-subject factor (context/occasion: Waiting Room, Gesell Room, Decompression Room). A priori power analysis (G*Power 3.1.9.4) indicated N=42 for medium effect (partial η²=0.06), α=0.05, power=0.95, 1 group, 3 measurements. Participants: N=42 minors (Spanish nationals) testifying as victims of violence at the Pilot Court (Las Palmas de Gran Canaria, Spain). Girls: 73.8% (n=31), age 4–17 (M=12.5, SD=3.4); Boys: 26.2% (n=11), age 7–17 (M=11.4, SD=3.1). Crime type: sexual assault 64.3%; other violence/abuse 35.7%. Perpetrators: father 23.8%, other relative 16.7%, mother 14.3%, non-family 45.2%. Inclusion: minors testifying in modern Gesell room. Exclusion: >18 years, not testifying in Gesell room. All assessed as fit to testify by court psychologist. Consent obtained from guardians; assent from children. Procedure and occasions:

• Waiting Room (before; T1): Explain objectives, verbal consent, device intro and calibration (age, weight, activity level, no identifiers). Researcher monitored stress from adjacent room, first measure after 5 minutes.
• Gesell Room (during; T2): Child-friendly design with CCTV; Achieving Best Evidence Interview by same interviewer with standard protocol. Stress measured 10 minutes after interview began while children freely recounted events/responded to open-ended questions.
• Decompression Room (after; T3–T6 resting measures at 3-min intervals during questionnaire completion; T7 final measure before leaving). Researcher cleared app cache post-session. Instruments:
• Anxiety: For <15 years, Spanish STAIC (State and Trait; 20 items each; KR-20: 0.91 state, 0.87 trait; percentile cutoffs: low 1–33, moderate 34–66, high ≥67). For ≥16 years, Spanish STAI (state and trait; reliability α: 0.964 state, 0.927 trait; percentiles: low 1–30, medium 31–70, high ≥71). Administered once in decompression room.
• Real-time stress: Huawei Band 6 with Huawei Health app; PPG-based HR and HRV. App algorithm uses HRV plus sex/age to compute standardized percentile stress score (ranges: low 1–29, medium-low 30–59, medium-high 60–79, high ≥80). Device selected for manual real-time measurement capability, low cost, portability, and prior validation/concordance studies. Data analysis:
• Reliability: Test–retest Pearson correlations among four resting decompression measures (T3–T6).
• Validity: Discriminant validity via correlations across contexts; convergent validity via correlations between stress and S/T-Anxiety. Percentile scores used for both stress and anxiety to align scales and manage distributional properties.
• Descriptives: Means, SDs, IQRs for stress percentiles by occasion; anxiety percentiles (state, trait).
• Group comparisons: Independent-samples t-tests by gender and crime type (sexual assault vs other) for anxiety and stress.
• Within-subject effects: Repeated-measures ANOVA (within factor: room/occasion), Shapiro–Wilk normality; Mauchly’s test for sphericity; Bonferroni post hoc comparisons. SPSS v28 used; α=0.05.

• Reliability/validity:
  • Test–retest: Decompression room resting stress measures (T3–T6) showed significant to highly significant bivariate correlations, indicating stability.
  • Discriminant: Waiting room stress did not significantly correlate with Gesell or Decompression; Gesell stress correlated positively and significantly with Decompression stress.
  • Convergent: S-Anxiety and T-Anxiety correlated positively with Waiting room stress (S: p<0.05; T: p<0.01) but not with Gesell or Decompression stress.
• Stress category distributions:
  • Waiting room: low 2.4%, medium-low 81.0%, medium-high 16.7%, high 0%.
  • Gesell room: low 9.5%, medium-low 66.7%, medium-high 16.7%, high 7.1% (only three children reached high, all here).
  • Decompression room: low 19.0%, medium-low 71.4%, medium-high 9.5%, high 0%.
• Descriptives (percentiles):
  • Anxiety: S-Anxiety mean 53.62 (IQR 56); T-Anxiety mean 55.12 (IQR 61).
  • Stress: Waiting 47.32 (IQR 12), Gesell 50.14 (IQR 20.32), Decompression 39.93 (IQR 16).
• Group differences:
  • No significant differences in anxiety or stress by gender or by type of crime. Victims of sexual assault were significantly older than victims of other violence; age did not correlate with stress; age correlated with T-Anxiety (r=0.37, p<0.05), not with S-Anxiety.
• Within-subject effects (ANOVA):
  • Significant effect of occasion on stress: F(2)=9.91, p<0.001, η²=0.195, observed power 0.981.
  • Bonferroni post hoc:
    • Gesell > Decompression: mean difference 10.214, SE 2.359, 95% CI [5.450, 14.978], p<0.001.
    • Waiting > Decompression: mean difference 7.399, SE 2.117, 95% CI [2.115, 12.683], p<0.05.
    • Waiting vs Gesell: not significant (mean difference −2.815, SE 2.607, 95% CI [−8.081, 2.450]). Overall, stress rose during testimony relative to post-test decompression but did not, on average, reach high or health-detrimental levels; anxiety levels were moderate.

Findings indicate that, within a child-friendly investigative court, children’s real-time physiological stress (HRV-derived) increases during testimony in the Gesell room and is similar to anticipatory levels in the waiting room, but declines significantly in the decompression room. Crucially, stress did not, on average, reach levels considered harmful to health, and only a small subset briefly reached high levels during testimony. This supports policies emphasizing adapted environments, procedures, and language to mitigate secondary victimization. The pattern suggests three phases: anticipatory stress before testimony, a peak during recounting of traumatic events, and relief afterward. Moderate anxiety scores and limited correlations between questionnaire-based anxiety and HRV-based stress (significant only for waiting room stress) align with models distinguishing state anxiety and acute physiological stress responses (Multidimensional Anxiety and Coping Theory). The results extend prior simulated-trial findings (higher stress in courtroom-like settings) to real victims in real-time within a child-friendly court, showing that such adaptations may keep stress within non-harmful ranges. The smart band measures demonstrated stability under resting conditions and discriminated between more and less stressful contexts, supporting HRV wearables’ utility. However, causality cannot be inferred from the design, and potential habituation or individual differences in stress reactivity may influence outcomes.

In this first real-time field study with child victims in a child-friendly investigative court, stress levels increased during testimony in the Gesell room but did not reach health-detrimental levels on average; stress was higher in the waiting and Gesell rooms than in the decompression room. Anxiety was generally moderate, with no significant differences by gender or crime type. These findings support European judicial policies advocating specialized, child-adapted courts and procedures to reduce stress and secondary victimization while preserving the child’s participation. Future research should include larger and more gender-balanced samples, incorporate control or comparison groups from non-adapted courts, measure distress explicitly, and assess state anxiety pre- and post-testimony to clarify causal effects and refine protective practices.

• Convergent validity with the Spielberger anxiety questionnaires was weaker than expected; anxiety and stress capture related but distinct constructs.
• No distress scale was administered.
• Modest sample size (N=42) and gender imbalance limit subgroup analyses and generalizability.
• No control/comparison group; causal attributions to the adapted environment cannot be made.
• Chronic stress was not assessed; focus was solely on acute stress.
• Prior consent and information about study aims may have influenced questionnaire responses (though less likely to affect objective HRV measures).
• State anxiety was measured only after testimony, not before/during, limiting temporal comparisons.